System and method for counting people flow

A statistical system and people flow technology, applied to neural learning methods, calculations, computer components, etc., can solve problems such as video recognition method failure, crowded carriages, and high overlap rate of human figures, so as to save resource expenses, reduce difficulty, and improve The effect of precision and recall

Active Publication Date: 2021-11-30
桂林海威科技股份有限公司
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In the existing technology, when the number of people in the car is small and the overlapping rate of human figures is low, the traditional barrel camera can be used as an accurate identification target; There will be irregular borders and completely occluded characters, which will lead to the failure of traditional video recognition methods

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • System and method for counting people flow
  • System and method for counting people flow

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0039] Such as figure 1 As shown, a people flow counting system includes:

[0040] Image collection module 1, described image collection module 1 is used for collecting the video data of flow of people in the enclosed area by fisheye camera, described video data is carried out frame extraction, obtains image set;

[0041] Model training module 2, described model training module 2 is used for importing image set, divides the image in image set into outer circle area and inner circle area; In the outer circle area of ​​image, mark out the facial features and the head of human body, obtain the first Human body marking data; using the first human body marking data to train the first human head detection model to obtain the first target detection model; marking the head of the human body in the inner circle area of ​​the image to obtain the second human body marking data; using the first human body marking data The second human head detection model is trained on the two human body...

Embodiment 2

[0054] Such as figure 2 Shown, a kind of people flow counting method comprises the following steps:

[0055] Collect the video data of the flow of people in the enclosed area through the fisheye camera, and frame the video data to obtain the image set;

[0056] Divide the images in the image set into the outer circle area and the inner circle area; mark the facial features and head of the human body in the outer circle area of ​​the image to obtain the first human body marking data; use the first human body marking data to detect the first head detection model Perform training to obtain the first target detection model; mark the head of the human body in the inner circle area of ​​the image to obtain the second human body marking data; use the second human body marking data to train the second human head detection model to obtain the second target detection model;

[0057] Divide the images in the image set to be recognized into the outer circle area and the inner circle ar...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention relates to a people flow statistics system and method, comprising an image acquisition module collecting video data of people flow in a closed area through a fisheye camera, and extracting frames from the video data to obtain an image set; a model training module importing the image set, Mark the facial features and head of the human body in the outer circle area, and obtain the human body marking data to train the first head detection model to obtain the first target detection model; mark the head of the human body in the inner circle area of ​​the image, and obtain the human body Mark the data to train the second head detection model to obtain the second target detection model; the human flow statistics module imports the image set to be recognized, imports the images into the first target detection model and the second target detection model, and infers the images to obtain People data in the current image; round up the moving average of the number of people in the image set to get the people flow data in the enclosed area. Compared with the prior art, the present invention can improve the efficiency and accuracy of people counting.

Description

technical field [0001] The present invention relates to the technical field of rail transit, in particular to a system and method for counting people flow. Background technique [0002] As an important basic data for the normal operation of equipment in closed spaces such as subway cars, elevators, and buses, the current number of people is an indispensable indicator in the fields of rail transit configuration, market decision-making, and safety precautions. The total number of people counting method based on fisheye imaging video image analysis technology has the advantages of high statistical accuracy and convenient implementation, so it is widely used. [0003] The traditional methods of counting the number of people in rail transit include: 1. Identify the specific number of people in the video by human eyes; 2. Count the number of people in the current car by entering and leaving the door; 3. Estimate the number of people in the current car by the weight of the car. ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/04G06N3/08G06V20/53G06V20/46G06V2201/07G06F18/214
Inventor 莫家佳周明资明祥张志斌覃琨王建卫
Owner 桂林海威科技股份有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap